cokit-cli 1.0.3 → 1.0.5

package/templates/repo/.github/prompts/ck-design-good.prompt.md

@@ -0,0 +1,36 @@
+---
+agent: 'agent'
+description: 'Create an immersive design'
+tools: ['search/codebase', 'web/fetch']
+---
+
+Think hard to plan & start working on these tasks follow the Orchestration Protocol, Core Responsibilities, Subagents Team and Development Rules:
+<tasks>${input}</tasks>
+
+## Required Skills (Priority Order)
+1. **`ui-ux-pro-max`** - Design intelligence database (ALWAYS ACTIVATE FIRST)
+2. **`frontend-design`** - Screenshot analysis and design replication
+
+**Ensure token efficiency while maintaining high quality.**
+
+## Workflow:
+1. **FIRST**: Run `ui-ux-pro-max` searches to gather design intelligence:
+   ```bash
+   python3  "<product-type>" --domain product
+   python3  "<style-keywords>" --domain style
+   python3  "<mood>" --domain typography
+   python3  "<industry>" --domain color
+   ```
+2. research about design style, trends, fonts, colors, border, spacing, elements' positions, etc.
+3. implement the design step by step based on the research.
+4. If user doesn't specify, create the design in pure HTML/CSS/JS.
+5. Report back to user with a summary of the changes and explain everything briefly, ask user to review the changes and approve them.
+6. If user approves the changes, update the `./docs/design-guidelines.md` docs if needed.
+
+## Important Notes:
+- **ALWAYS REMEBER that you have the skills of a top-tier UI/UX Designer who won a lot of awards on Dribbble, Behance, Awwwards, Mobbin, TheFWA.**
+- Remember that you have the capability to generate images, videos, edit images, etc. with `ai-multimodal` skills for image generation. Use them to create the design with real assets.
+- Always review, analyze and double check the generated assets with `ai-multimodal` skills to verify quality.
+- Use removal background tools to remove background from generated assets if needed.
+- Create storytelling designs, immersive 3D experiences, micro-interactions, and interactive interfaces.
+- Maintain and update `./docs/design-guidelines.md` docs if needed.

package/templates/repo/.github/prompts/ck-design-screenshot.prompt.md

@@ -0,0 +1,35 @@
+---
+agent: 'agent'
+description: 'Create a design based on screenshot'
+tools: ['search/codebase', 'web/fetch']
+---
+
+Think hard to plan & start designing follow exactly this screenshot:
+<screenshot>${input}</screenshot>
+
+## Required Skills (Priority Order)
+1. **`ui-ux-pro-max`** - Design intelligence database (ALWAYS ACTIVATE FIRST)
+2. **`frontend-design`** - Screenshot analysis and replication
+
+**Ensure token efficiency while maintaining high quality.**
+
+## Workflow:
+1. )
+   - **IMPORTANT:** Try to predict the font name (Google Fonts) and font size in the given screenshot, don't just use Inter or Poppins.
+2. create a design plan following the progressive disclosure structure so the final result matches the screenshot:
+   - Create a directory using naming pattern from `## Naming` section.
+   - Save the overview access point at `plan.md`, keep it generic, under 80 lines, and list each phase with status/progress and links.
+   - For each phase, add `phase-XX-phase-name.md` files containing sections (Context links, Overview with date/priority/statuses, Key Insights, Requirements, Architecture, Related code files, Implementation Steps, Todo list, Success Criteria, Risk Assessment, Security Considerations, Next steps).
+   - Keep every research markdown report concise (≤150 lines) while covering all requested topics and citations.
+3. Then implement the plan step by step.
+4. If user doesn't specify, create the design in pure HTML/CSS/JS.
+5. Report back to user with a summary of the changes and explain everything briefly, ask user to review the changes and approve them.
+6. If user approves the changes, update the `./docs/design-guidelines.md` docs if needed.
+
+## Important Notes:
+- **ALWAYS REMEBER that you have the skills of a top-tier UI/UX Designer who won a lot of awards on Dribbble, Behance, Awwwards, Mobbin, TheFWA.**
+- Remember that you have the capability to generate images, videos, edit images, etc. with ai-multimodal skill for image generation. Use them to create the design with real assets.
+- Always review, analyze and double check the generated assets with ai-multimodal skill to verify quality.
+- Use removal background tools to remove background from generated assets if needed.
+- Create storytelling designs, immersive 3D experiences, micro-interactions, and interactive interfaces.
+- Maintain and update `./docs/design-guidelines.md` docs if needed.

package/templates/repo/.github/prompts/ck-design-video.prompt.md

@@ -0,0 +1,35 @@
+---
+agent: 'agent'
+description: 'Create a design based on video'
+tools: ['search/codebase', 'web/fetch']
+---
+
+Think hard to plan & start designing follow exactly this video:
+<video>${input}</video>
+
+## Required Skills (Priority Order)
+1. **`ui-ux-pro-max`** - Design intelligence database (ALWAYS ACTIVATE FIRST)
+2. **`frontend-design`** - Video analysis and replication
+
+**Ensure token efficiency while maintaining high quality.**
+
+## Workflow:
+1. 
+   - **IMPORTANT:** Try to predict the font name (Google Fonts) and font size in the given video, don't just use Inter or Poppins.
+2. create a design plan following the progressive disclosure structure so the final result matches the video:
+   - Create a directory using naming pattern from `## Naming` section.
+   - Save the overview access point at `plan.md`, keep it generic, under 80 lines, and list each phase with status/progress and links.
+   - For each phase, add `phase-XX-phase-name.md` files containing sections (Context links, Overview with date/priority/statuses, Key Insights, Requirements, Architecture, Related code files, Implementation Steps, Todo list, Success Criteria, Risk Assessment, Security Considerations, Next steps).
+   - Keep every research markdown report concise (≤150 lines) while covering all requested topics and citations.
+3. Then implement the plan step by step.
+4. If user doesn't specify, create the design in pure HTML/CSS/JS.
+5. Report back to user with a summary of the changes and explain everything briefly, ask user to review the changes and approve them.
+6. If user approves the changes, update the `./docs/design-guidelines.md` docs if needed.
+
+## Important Notes:
+- **ALWAYS REMEBER that you have the skills of a top-tier UI/UX Designer who won a lot of awards on Dribbble, Behance, Awwwards, Mobbin, TheFWA.**
+- Remember that you have the capability to generate images, videos, edit images, etc. with ai-multimodal skill for image generation. Use them to create the design with real assets.
+- Always review, analyze and double check the generated assets with ai-multimodal skill to verify quality.
+- Use removal background tools to remove background from generated assets if needed.
+- Create storytelling designs, immersive 3D experiences, micro-interactions, and interactive interfaces.
+- Maintain and update `./docs/design-guidelines.md` docs if needed.

package/templates/repo/.github/prompts/ck-docs-init.prompt.md

@@ -0,0 +1,27 @@
+---
+agent: 'agent'
+description: 'Analyze the codebase and create initial documentation'
+tools: ['search/codebase', 'search/changes']
+---
+
+## Phase 1: Parallel Codebase Scouting
+
+**You (main agent) must 
+
+1. Run `ls -la` to identify actual project directories
+2. Spawn 2-4 `scout-external` (preferred, uses Gemini 2M context) or `scout` (fallback) via 
+3. Target directories **that actually exist** - adapt to project structure, don't hardcode paths
+4. Merge scout results into context summary
+
+## Phase 2: Documentation Creation (docs-manager Agent)
+
+Pass the gathered file list to `docs-manager` agent to create initial documentation:
+- `docs/project-overview-pdr.md`: Project overview and PDR (Product Development Requirements)
+- `./docs/codebase-summary.md`: Codebase summary
+- ``: Codebase structure and code standards
+- `docs/system-architecture.md`: System architecture
+- Update `README.md` with initial documentation (keep it under 300 lines)
+
+Use `docs/` directory as the source of truth for documentation.
+
+**IMPORTANT**: **Do not** start implementing.

package/templates/repo/.github/prompts/ck-docs-summarize.prompt.md

@@ -0,0 +1,23 @@
+---
+agent: 'agent'
+description: 'Analyze the codebase and update documentation'
+tools: ['search/codebase', 'search/changes']
+---
+
+Analyze the codebase based on `./docs/codebase-summary.md` and respond with a summary report.
+
+## Arguments:
+$1: Focused topics (default: all)
+$2: Should scan codebase (`Boolean`, default: `false`)
+
+## Focused Topics:
+<focused_topics>$1</focused_topics>
+
+## Should Scan Codebase:
+<should_scan_codebase>$2</should_scan_codebase>
+
+## Important:
+- Use `docs/` directory as the source of truth for documentation.
+- Do not scan the entire codebase unless the user explicitly requests it.
+
+**IMPORTANT**: **Do not** start implementing.

package/templates/repo/.github/prompts/ck-docs-update.prompt.md

@@ -0,0 +1,36 @@
+---
+agent: 'agent'
+description: 'Analyze the codebase and update documentation'
+tools: ['search/codebase', 'search/changes']
+---
+
+## Phase 1: Parallel Codebase Scouting
+
+**You (main agent) must 
+
+1. Run `ls -la` to identify actual project directories
+2. Spawn 2-4 `scout-external` (preferred, uses Gemini 2M context) or `scout` (fallback) via 
+3. Target directories **that actually exist** - adapt to project structure, don't hardcode paths
+4. Merge scout results into context summary
+
+## Phase 2: Documentation Update (docs-manager Agent)
+
+Pass the gathered file list to `docs-manager` agent to update documentation:
+- `README.md`: Update README (keep it under 300 lines)
+- `docs/project-overview-pdr.md`: Update project overview and PDR (Product Development Requirements)
+- `./docs/codebase-summary.md`: Update codebase summary
+- ``: Update codebase structure and code standards
+- `docs/system-architecture.md`: Update system architecture
+- `docs/project-roadmap.md`: Update project roadmap
+- `docs/deployment-guide.md` [optional]: Update deployment guide
+- `docs/design-guidelines.md` [optional]: Update design guidelines
+
+## Additional requests
+<additional_requests>
+  ${input}
+</additional_requests>
+
+## Important
+- Use `docs/` directory as the source of truth for documentation.
+
+**IMPORTANT**: **Do not** start implementing.

package/templates/repo/.github/prompts/ck-fix-ci.prompt.md

@@ -0,0 +1,18 @@
+---
+agent: 'agent'
+description: 'Analyze Github Actions logs and fix issues'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'read/terminalLastCommand', 'web/githubRepo']
+---
+
+## Github Actions URL
+<url>${input}</url>
+
+## Workflow
+1. read the github actions logs with `gh` command, analyze and find the root cause of the issues and .
+2. Start implementing the fix based the reports and solutions.
+3. test the fix and make sure it works, .
+4. If there are issues or failed tests, repeat from step 2.
+5. After finishing, respond back to user with a summary of the changes and explain everything briefly, guide user to get started and suggest the next steps.
+
+## Notes
+- If `gh` command is not available, instruct the user to install and authorize GitHub CLI first.

package/templates/repo/.github/prompts/ck-fix-fast.prompt.md

@@ -0,0 +1,18 @@
+---
+agent: 'agent'
+description: 'Analyze and fix small issues [FAST]'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'read/terminalLastCommand']
+---
+
+## Mission
+**Think hard** to analyze and fix these issues:
+<issues>${input}</issues>
+
+## Workflow
+1. If the user provides a screenshots or videos, 
+2. find the root cause of the issues and .
+3. Activate `debugging` skills and `problem-solving` skills to tackle the issues.
+4. Start implementing the fix based the reports and solutions.
+5. test the fix and make sure it works, .
+6. If there are issues or failed tests, repeat from step 2.
+7. After finishing, respond back to user with a summary of the changes and explain everything briefly, guide user to get started and suggest the next steps.

package/templates/repo/.github/prompts/ck-fix-hard.prompt.md

@@ -0,0 +1,36 @@
+---
+agent: 'agent'
+description: 'Use planner to plan and fix hard issues'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'read/terminalLastCommand', 'web/fetch']
+---
+
+**Ultrathink** to plan & start fixing these issues follow the Orchestration Protocol, Core Responsibilities, Subagents Team and Development Rules:
+<issues>${input}</issues>
+
+## Workflow:
+
+If the user provides a screenshots or videos, 
+
+### Fullfill the request
+**Question Everything**: Use `AskUserQuestion` tool to ask probing questions to fully understand the user's request, constraints, and true objectives. Don't assume - clarify until you're 100% certain.
+
+* If you have any questions, use `AskUserQuestion` tool to ask the user to clarify them.
+* Ask 1 question at a time, wait for the user to answer before moving to the next question.
+* If you don't have any questions, start the next step.
+
+### Fix the issue
+
+1. find the root cause of the issues and .
+2. research quickly about the root causes on the internet (if needed) and .
+3. create an implementation plan based on the reports, .
+4. Then use  SlashCommand to implement the plan step by step.
+5. Final Report:
+  * Report back to user with a summary of the changes and explain everything briefly, guide user to get started and suggest the next steps.
+  * Ask the user if they want to commit and push to git repository, if yes, commit and push to git repository.
+  - **IMPORTANT:** Sacrifice grammar for the sake of concision when writing reports.
+  - **IMPORTANT:** In reports, list any unresolved questions at the end, if any.
+
+**REMEMBER**:
+- You can always generate images with `ai-multimodal` skills on the fly for visual assets.
+- You always read and analyze the generated assets with `ai-multimodal` skills to verify they meet requirements.
+- For image editing (removing background, adjusting, cropping), use `ImageMagick` skill or similar tools as needed.

package/templates/repo/.github/prompts/ck-fix-logs.prompt.md

@@ -0,0 +1,27 @@
+---
+agent: 'agent'
+description: 'Analyze logs and fix issues'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'read/terminalLastCommand']
+---
+
+**IMPORTANT:** Activate `debugging` and `problem-solving` skills.
+
+## Mission
+<issue>${input}</issue>
+
+## Workflow
+1. Check if `./logs.txt` exists:
+   - If missing, set up permanent log piping in project's script config (`package.json`, `Makefile`, `pyproject.toml`, etc.):
+     - **Bash/Unix**: append `2>&1 | tee logs.txt`
+     - **PowerShell**: append `*>&1 | Tee-Object logs.txt`
+   - Run the command to generate logs
+2. analyze `./logs.txt` and find root causes:
+   - Use `Grep` with `head_limit: 30` to read only last 30 lines (avoid loading entire file)
+   - If insufficient context, increase `head_limit` as needed
+3. analyze the codebase and find the exact location of the issues, .
+4. create an implementation plan based on the reports, .
+5. Start implementing the fix based the reports and solutions.
+6. test the fix and make sure it works, .
+7. quickly review the code changes and make sure it meets requirements, .
+8. If there are issues or failed tests, repeat from step 3.
+9. After finishing, respond back to user with a summary of the changes and explain everything briefly, guide user to get started and suggest the next steps.

package/templates/repo/.github/prompts/ck-fix-parallel.prompt.md

@@ -0,0 +1,54 @@
+---
+agent: 'agent'
+description: 'Analyze & fix issues with parallel fullstack-developer agents'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'read/terminalLastCommand']
+---
+
+**Ultrathink parallel** to fix: <issues>${input}</issues>
+
+**IMPORTANT:** Activate needed skills. Ensure token efficiency. Sacrifice grammar for concision.
+
+## Workflow
+
+### 1. Issue Analysis
+- analyze root causes
+- Use  to find related files
+- Categorize issues by scope/area (frontend, backend, auth, payments, etc.)
+- Identify dependencies between issues
+
+### 2. Parallel Fix Planning
+- Trigger :parallel <detailed-fix-instructions>` for parallel-executable fix plan
+- Wait for plan with dependency graph, execution strategy, file ownership matrix
+- Group independent fixes for parallel execution
+- Sequential fixes for dependent issues
+
+### 3. Parallel Fix Implementation
+- Read `plan.md` for dependency graph
+- Launch multiple `fullstack-developer` agents  for independent fixes
+  - Example: "Fix auth + Fix payments + Fix UI" → launch 3 agents simultaneously
+  - Pass phase file path: `{plan-dir}/phase-XX-*.md`
+  - Include environment info
+- Wait for all parallel fixes complete before dependent fixes
+- Sequential fixes: launch one agent at a time
+
+### 4. Testing
+- NO fake data/mocks/cheats
+- Verify all issues resolved
+- If fail: use `debugger`, fix, repeat
+
+### 5. Code Review
+- Use `code-reviewer` for all changes
+- Verify fixes don't introduce regressions
+- If critical issues: fix, retest
+
+### 6. Project Management & Docs
+- If approved: use `project-manager` + `docs-manager` 
+- Update plan files, docs, roadmap
+- If rejected: fix and repeat
+
+### 7. Final Report
+- Summary of all fixes from parallel phases
+- Verification status per issue
+- Ask to commit (use `git-manager` if yes)
+
+**Example:** Fix 1 (auth) + Fix 2 (payments) + Fix 3 (UI) → Launch 3 fullstack-developer agents → Wait → Fix 4 (integration) sequential

package/templates/repo/.github/prompts/ck-fix-test.prompt.md

@@ -0,0 +1,19 @@
+---
+agent: 'agent'
+description: 'Run test suite and fix issues'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'read/terminalLastCommand']
+---
+
+## Reported Issues:
+<issues>${input}</issues>
+
+## Workflow:
+1. compile the code and fix all syntax errors if any.
+2. run the tests and .
+3. If there are issues or failed tests, find the root cause of the issues, .
+4. create an implementation plan based on the reports, .
+5. Use main agent to implement the plan step by step.
+6. test the fix and make sure it works, .
+6. quickly review the code changes and make sure it meets requirements, .
+7. If there are issues or failed tests, repeat from step 2.
+8. After finishing, respond back to user with a summary of the changes and explain everything briefly, guide user to get started and suggest the next steps.

package/templates/repo/.github/prompts/ck-fix-types.prompt.md

@@ -0,0 +1,11 @@
+---
+agent: 'agent'
+description: 'Fix type errors'
+tools: ['search/codebase', 'read/problems', 'read/terminalLastCommand']
+---
+
+Run `bun run typecheck` or `tsc` or `npx tsc` and fix all type errors.
+
+## Rules
+- Fix all of type errors and repeat the process until there are no more type errors.
+- Do not use `any` just to pass the type check.

package/templates/repo/.github/prompts/ck-fix-ui.prompt.md

@@ -0,0 +1,49 @@
+---
+agent: 'agent'
+description: 'Analyze and fix UI issues'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'web/fetch']
+---
+
+## Required Skills (Priority Order)
+1. **`ui-ux-pro-max`** - Design intelligence database (ALWAYS ACTIVATE FIRST)
+2. **`aesthetic`** - Design principles
+3. **`frontend-design`** - Implementation patterns
+
+Read and analyze `./docs/design-guidelines.md` then fix the following issues:
+<issue>${input}</issue>
+
+## Workflow
+**FIRST**: Run `ui-ux-pro-max` searches to understand context and common issues:
+```bash
+python3  "<product-type>" --domain product
+python3  "<style-keywords>" --domain style
+python3  "accessibility" --domain ux
+python3  "z-index animation" --domain ux
+```
+
+If the user provides a screenshots or videos, 
+
+1. implement the fix step by step.
+2. Use screenshot capture tools along with `ai-multimodal` skill to take screenshots of the implemented fix (at the exact parent container, don't take screenshot of the whole page) and use the appropriate Gemini analysis skills (`ai-multimodal`, `video-analysis`, or `document-extraction`) to analyze those outputs so the result matches the design guideline and addresses all issues.
+  - If the issues are not addressed, repeat the process until all issues are addressed.
+3. 
+4. test the fix and compile the code to make sure it works, .
+  - If there are issues or failed tests, ask main agent to fix all of them and repeat the process until all tests pass.
+5. Project Management & Documentation:
+  **If user approves the changes:** Use `project-manager` and s  to update the project progress and documentation:
+    * update the project progress and task status in the given plan file.
+    * update the docs in `./docs` directory if needed.
+    * create a project roadmap at `./docs/project-roadmap.md` file.
+    * **IMPORTANT:** Sacrifice grammar for the sake of concision when writing outputs.
+  **If user rejects the changes:** Ask user to explain the issues and ask main agent to fix all of them and repeat the process.
+6. Final Report:
+  * Report back to user with a summary of the changes and explain everything briefly, guide user to get started and suggest the next steps.
+  * Ask the user if they want to commit and push to git repository, if yes, commit and push to git repository.
+  * **IMPORTANT:** Sacrifice grammar for the sake of concision when writing reports.
+  * **IMPORTANT:** In reports, list any unresolved questions at the end, if any.
+
+**REMEMBER**:
+- You can always generate images with `ai-multimodal` skill on the fly for visual assets.
+- You always read and analyze the generated assets with `ai-multimodal` skill to verify they meet requirements.
+- For image editing (removing background, adjusting, cropping), use `ImageMagick` skill or similar tools as needed.
+- **IMPORTANT:** 

package/templates/repo/.github/prompts/ck-fix.prompt.md

@@ -1,24 +1,44 @@
 ---
-mode: agent
-description: Debug and fix code issues systematically
+agent: 'agent'
+description: 'Analyze and fix issues [INTELLIGENT ROUTING]'
+tools: ['search/codebase', 'search/changes', 'read/problems', 'read/terminalLastCommand']
 ---
 
-# Fix - Systematic Debugging
+**Analyze issues and route to specialized fix command:**
+<issues>${input}</issues>
 
-You are a debugging expert. Help the user fix their issue systematically.
+## Decision Tree
 
-## Approach
+**1. Check for existing plan:**
+- If markdown plan exists →  <path-to-plan>`
 
-1. **Understand the problem** - What's the expected vs actual behavior?
-2. **Reproduce** - Can you identify where the issue occurs?
-3. **Root cause** - Find the actual cause, not just symptoms
-4. **Fix** - Make the minimal change needed
-5. **Verify** - Confirm the fix works and doesn't break anything else
+**2. Route by issue type:**
 
-## Guidelines
+**A) Type Errors** (keywords: type, typescript, tsc, type error)
+→ `/fix:types`
 
-- Ask clarifying questions if needed
-- Check logs and error messages
-- Don't guess - investigate first
-- Explain what you find
-- Keep fixes focused and minimal
+**B) UI/UX Issues** (keywords: ui, ux, design, layout, style, visual, button, component, css, responsive)
+→ `/fix:ui <detailed-description>`
+
+**C) CI/CD Issues** (keywords: github actions, pipeline, ci/cd, workflow, deployment, build failed)
+→ `/fix:ci <github-actions-url-or-description>`
+
+**D) Test Failures** (keywords: test, spec, jest, vitest, failing test, test suite)
+→ `/fix:test <detailed-description>`
+
+**E) Log Analysis** (keywords: logs, error logs, log file, stack trace)
+→ `/fix:logs <detailed-description>`
+
+**F) Multiple Independent Issues** (2+ unrelated issues in different areas)
+→ `/fix:parallel <detailed-description>`
+
+**G) Complex Issues** (keywords: complex, architecture, refactor, major, system-wide, multiple components)
+→ `/fix:hard <detailed-description>`
+
+**H) Simple/Quick Fixes** (default: small bug, single file, straightforward)
+→ `/fix:fast <detailed-description>`
+
+## Notes
+- `detailed-description` = enhanced prompt describing issue in detail
+- If unclear, ask user for clarification before routing
+- Can combine routes: e.g., multiple type errors + UI issue → `/fix:parallel`

package/templates/repo/.github/prompts/ck-git-cm.prompt.md

@@ -0,0 +1,8 @@
+---
+agent: 'agent'
+description: 'Stage all files and create a commit.'
+tools: ['search/changes', 'web/githubRepo']
+---
+
+stage all files and create a commit.
+**IMPORTANT: DO NOT push the changes to remote repository**

package/templates/repo/.github/prompts/ck-git-cp.prompt.md

@@ -0,0 +1,7 @@
+---
+agent: 'agent'
+description: 'Stage, commit and push all code in the current branch'
+tools: ['search/changes', 'web/githubRepo']
+---
+
+stage all files, create a meaningful commit based on the changes and push to remote repository.

package/templates/repo/.github/prompts/ck-git-merge.prompt.md

@@ -0,0 +1,41 @@
+---
+agent: 'agent'
+description: '⚠️ Merge code from one branch to another'
+tools: ['search/changes', 'web/githubRepo']
+---
+
+## Variables
+
+TO_BRANCH: $1 (defaults to `main`)
+FROM_BRANCH: $2 (defaults to current branch)
+
+## Workflow
+
+### Step 1: Sync with remote (CRITICAL)
+```bash
+git fetch origin
+git checkout {TO_BRANCH}
+git pull origin {TO_BRANCH}
+```
+
+### Step 2: Merge from REMOTE tracking branch
+```bash
+# Use origin/{FROM_BRANCH} to merge remote state, not local WIP
+git merge origin/{FROM_BRANCH} --no-ff -m "merge: {FROM_BRANCH} into {TO_BRANCH}"
+```
+
+**Why `origin/{FROM_BRANCH}`:** Ensures merging only committed+pushed changes, not local uncommitted work.
+
+### Step 3: Resolve conflicts if any
+- If conflicts exist, resolve them manually
+- After resolution: `git add . && git commit`
+
+### Step 4: Push merged result
+```bash
+git push origin {TO_BRANCH}
+```
+
+## Notes
+- If `gh` command is not available, instruct the user to install and authorize GitHub CLI first.
+- If you need more clarifications, use `AskUserQuestion` tool to ask the user for more details.
+- Always fetch and pull latest remote state before merging to avoid stale conflicts.

package/templates/repo/.github/prompts/ck-git-pr.prompt.md

@@ -0,0 +1,49 @@
+---
+agent: 'agent'
+description: 'Create a pull request'
+tools: ['search/changes', 'web/githubRepo']
+---
+
+## Variables
+
+TO_BRANCH: $1 (defaults to `main`)
+FROM_BRANCH: $2 (defaults to current branch)
+
+## Workflow
+
+### Step 1: Ensure remote is synced
+```bash
+git fetch origin
+git push -u origin HEAD  # Push current branch if not pushed
+```
+
+### Step 2: Analyze REMOTE diff (CRITICAL)
+**IMPORTANT:** Always compare REMOTE branches, not local:
+```bash
+# Get commits between remote branches (what PR will actually contain)
+git log origin/{TO_BRANCH}...origin/{FROM_BRANCH} --oneline
+
+# Get file diff between remote branches
+git diff origin/{TO_BRANCH}...origin/{FROM_BRANCH} --stat
+git diff origin/{TO_BRANCH}...origin/{FROM_BRANCH}
+```
+
+**DO NOT use:**
+- `git diff {TO_BRANCH}...HEAD` (includes unpushed local changes)
+- `git diff --cached` (staged local changes)
+- `git status` (local working tree state)
+
+### Step 3: Generate PR content from remote diff
+Based on the REMOTE diff analysis:
+- **Title:** Conventional commit format from the primary change (no version/release numbers)
+- **Body:** Summary of changes that exist ON REMOTE, not local WIP
+
+### Step 4: Create PR
+```bash
+gh pr create --base {TO_BRANCH} --head {FROM_BRANCH} --title "..." --body "..."
+```
+
+## Notes
+- If `gh` command is not available, instruct the user to install and authorize GitHub CLI first.
+- If local has unpushed commits, push first before analyzing diff.
+- PR content must reflect REMOTE state since PRs are based on remote branches.